When I want to get dense layer from the network when doing train, got the
following error. How do fix this?
My input:
self.input_var = T.tensor3('inputs')
self.target_var = T.ivector('targets')
get out of dense layer:
self.util_add_dense(self.dense1_sz)
self.dense1_out = lasagne.layers.get_output(self.network,
deterministic=True)
self.dense1 = theano.function([self.input_var], [self.dense1_out])
in train function:
dense1 = networkrel.dense1(inputs)
Traceback (most recent call last):
File "relational_network.py", line 211, in <module>
semantic_ae.network_train()
File "relational_network.py", line 203, in network_train
utility_train(self, self.training_inputs, self.training_targets,
self.test_inputs, self.test_targets)
File "/home/relational_network_utilities.py", line 91, in utility_train
dense = networkrel.dense1(inputs, targets)
File
"/root/miniconda2/lib/python2.7/site-packages/theano/compile/function_module.py",
line 813, in __call__
allow_downcast=s.allow_downcast)
File
"/root/miniconda2/lib/python2.7/site-packages/theano/tensor/type.py", line
140, in filter
raise TypeError(err_msg)
TypeError: Bad input argument to theano function with name
"relational_network.py:113" at index 0 (0-based).
Backtrace when that variable is created:
File "relational_network.py", line 209, in <module>
semantic_ae.network_build_layers_v1()
File "relational_network.py", line 104, in network_build_layers_v1
self.network_inp_prepare()
File "relational_network.py", line 89, in network_inp_prepare
self.input_var = T.tensor3('inputs')
TensorType(float32, 3D) cannot store a value of dtype float64 without
risking loss of precision. If you do not mind this loss, you can: 1)
explicitly cast your data to float32, or 2) set "allow_input_downcast=True"
when calling "function". Value: "array([[[0. , 0. , 0.
, ..., 0. ,
--
---
You received this message because you are subscribed to the Google Groups
"theano-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to theano-users+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/theano-users/dda37f68-6849-49f3-9d5d-14d08175905f%40googlegroups.com.
